Evaluación y Rehabilitación de Estructuras

introduction

Reduced Goal Directed Behaviour (GDB) is increasingly being recognised as an important feature of PD. Although older descriptions of a GDB impairment in this disease are available (see e.g. Charcot, 1880) only recently have systematic

investigations been performed into the aspects of the disease described as apathy. The first, by Starkstein et al. (1992) used a customised version o f the Apathy Evaluation Scale (AES) of Marin et al. (1991). This 14-point questionnaire was completed by 50 PD patients visiting a neurology clinic. The frequencies o f individual apathy scores indicated a bimodal distribution. It was estimated that 42% of the sample had apathy based on this distribution. A later report used the Neuropsychiatrie Inventory (NPI) of Cummings et al. (1994). This was administered to the carers of 40 PD patients during pre-operative assessment for pallidotomy surgery. It was found that 33% of the sample could be described as apathetic (Levy et al., 1998). The only other research to date, also using the NPI, was conducted on a community sample of 139 PD patients. This

produced a much smaller estimate of the prevalence of apathy in PD at 16.5% (Aarsland et al., 1999).

The discrepancy in estimates of prevalence probably reflects sampling effects. For example, it is highly likely that the pre-pallidotomy patients were, on average, more advanced cases than those reported in the community or outpatient samples. However, a major flaw in each o f these studies is the lack of a baseline of apathy scores in a non- neurological population. Starkstein et al. (1992), using the modified AES, assumed that

bimodality o f the distribution indicated those with and those without apathy. In the two studies that used the NPI, it was implicitly assumed that apathy is a pathological

response not present in patients without neurological dysfunction. Therefore, in none of the above reports was the distribution of scores in normal samples considered. This has two negative implications. First, it is impossible to estimate accurately how the PD patients differ from normal populations in terms o f apathy. Second, it is impossible to say whether apathy is a secondary social consequence of the chronic disease process, a distinct symptom reflecting the underlying pathology, or a non-specific physiological response to chronic illness. This latter possibility is raised by the observation that, in animals, illness reliably produces a range of behavioural changes including a loss of interest for daily activities (Aubert, 1999). This is likely to be a physiological rather than a behavioural adaptation, a key factor that produces a GDB change in sick animals is the level of cytokines in the body (Anisman, Kokkinidis, Borowski, & Merali, 1998). Therefore, specific physiological changes can cause reduced GDB. In PD patients, this may be an alternative explanation to apathy being a psychological reaction to physical impairment.

A further question is ‘what factors are associated with apathy in PD’? Many of these issues were discussed in Chapters 1 and 2 and so will only briefly be recapitulated here. The relationship between depression and apathy is important. Apathy is often not distinguished from depression in PD as a separate clinical entity (see e.g. Lieberman,

1998). In one review depression without sadness (sometimes called ‘masked depression’), has been described in older adults (Gallo & Rabins, 1999). It is

controversial whether depression should be diagnosed in the absence of negative mood, indeed emotional distress is crucial to many definitions of depression (see e.g.

distress then the state more readily resembles apathy as defined in Chapter 1. The problem arises due to symptoms of depression that overlap with the construct of apathy. Intuitively there are common aspects to each, and indeed depression has been described as showing signs o f ‘sick-will’ (Ingvar, 1994). Clearly, the relationship between

depression and apathy is worthy of study.

The conclusions from what little research has been performed into the relationship between apathy and depression in PD are unclear. Starkstein et al. (1992) found that the majority of patients they described as apathetic were also depressed. While 42% of PD patients were described as apathetic, only 12% showed no signs of depression. This seems to indicate that there is a close co-morbidity of apathy and depression in PD. Levy et al. (1998) compared 154 patients with neurodegenerative disease, composed of patients with Alzheimer’s disease, fronto-temporal dementia, progressive supranuclear palsy, Huntington’s disease and PD. They failed to find a correlation between apathy and depression in this combined group. When only the scores o f the 40 PD patients were compared a significant but small (r=.34) association between apathy and depression was found. The most recent study by Aarsland et al. (1999) also found a significant relationship between apathy and depression. O f the 23 PD patients in the sample identified as having apathy, only six showed no signs o f depression.

It is already well known that anxiety is a common symptom in PD patients (Siemers et al., 1993). However, it is not usually associated with apathy. Disinterest, a feature of apathy, seems to be opposite to many common descriptions of anxiety that emphasise increased attention to stimuli or ‘hypervigilance’ (Wenzel & Holt, 1999). Empirical work has also tended to support the independence of anxiety and apathy. For example, Starkstein et al. (1992) failed to find a relationship between apathy and anxiety in their

PD sample. However, Arsland et al. (1999) used a factor analysis approach and found that in PD patients apathy and anxiety scores on the NPI were correlated and clustered into one factor. Furthermore, it has been shown that in normal samples low anxiety is linked to 'behavioural activation' (see e.g. Pickering et al., 1997). Therefore, raised anxiety may be related to apathy in PD patients. This hypothesis is tested below.

Other mood variables are also worthy o f consideration. Anhedonia, the inability to experience pleasure, is a concept closely related to apathy. It has been stated that

“anhedonia borders upon a number o f other constructs; these include diminution o f interest, reactivity o f mood, flattening o f affect, apathy and anergia ” (Snaith, 1993 p 957).

In comparative research, the dopamine system, particularly projections between the ventral tegmental area and the nucleus accumbens, has been identified as being closely associated with reward (Meredith & Totterdell, 1999). As discussed in Chapter 1, the dopamine system is also impaired in PD and so it could be hypothesised that the processing of reward is hindered in patients with PD. There is some evidence for this from PET studies. For example, it has been found that in a search task in which

financial rewards were manipulated, normal controls showed activation in the substantia nigra, nucleus accumbens and prefrontal cortex. This activation was absent in PD patients (Goerent et al., 1999). However, there are no direct observations of anhedonia in PD. The only minor exceptions being a psychophysics study o f taste preferences that found PD patients preferred higher sucrose concentrations than lower concentrations in a way that differed from the control subjects. The authors suggested that this might reflect impairment of the ‘hedonic’ value of the sensation (Travers et al., 1993). The other is a description of the reduced ability of a psychostimulant (methylphenidate) to

improve positive mood in PD patients compared to controls (Persico, Reich,

Henningfield, Kuhar, & Uhl, 1998). Consequently, although reward processing seems susceptible to the physiological changes seen in PD, the presence o f anhedonia has not been confirmed. It is therefore of interest to assess whether anhedonia is present in PD patients and its possible relationship to apathy.

A related issue is the question of personality in PD patients. A particular ‘parkinsonian personality’, characterised by inflexibility, moral rigidity and introversion has often been reported (Poewe et al., 1990). This is considered to be directly related to the dopaminergic abnormalities in PD brains (Menza, Mark, Burn, & Brooks, 1995). Indeed, dopamine is considered an important substrate of extraversion, a trait that is defined in terms of motivation, impulsivity and agency (Depue & Collins, 1999). Therefore both the specific personality and apathy described independently in PD may be different aspects of the same phenomenon. Alternatively, the occurrence of apathy and the parkinsonian personality may be relatively independent manifestations of physiological alterations in PD patients. At present, there is no clear way of distinguishing between these two hypotheses.

Cognition has been studied extensively in PD. Although the most commonly observed cognitive profile in PD patients is an executive disorder (Brown & Marsden, 1990), a range of other cognitive impairments has also been described. These include

abnormalities in memory (Hugdahl et al., 1993), visuo-spatial abilities (Cronin-Golomb & Braun, 1997), visual attention (Lieb et al., 1999) and semantic processing (Spicer et al., 1994). In several studies, executive skills have been shown to be impaired in apathetic patients with non-parkinsonian disorders such as Alzheimer’s disease (Kuzis et al., 1999) and HIV infection (Castellon et al., 1998). The only study that has looked

at cognition and apathy specifically in PD has also found that apathy is associated with executive dysfunction, but non-executive skills were not fully assessed (Starkstein et al., 1992). Therefore, it is unclear to what extent executive rather than non-executive cognitive skills are associated with apathy.

There are several questions that need to be answered, regarding apathy in PD. These include: 1) Is apathy a reaction to disability or a direct symptom of PD? 2) Are PD patients with apathy cognitively impaired? If so, what broad domains of function are implicated? 3) What is the relationship between apathy and other psychopathologies such as depression, anxiety and anhedonia in PD? 4) Are the parkinsonian personality and apathy in PD descriptions of the same phenomenon?

Method

In order to answer the questions stated above, 66 patients with either PD or

osteoarthritis (OA) were visited in their homes and given a wide range of assessments of disease characteristics, cognitive skills, psychiatric status and personality. OA patients were included as they were considered a suitable control group for the patients with PD. OA, like PD is a chronic, progressive, presently incurable illness that results in

difficulty or inability to perform day-to-day tasks with usual diagnosis in later life (Felson et al., 2000). Furthermore, PD and OA are similar in that once diagnosed, sufferers know that the disease will be part of the rest of their lives. There is therefore, no expected 'length of illness'. The use of a medical control group such as this is employed when it is necessary to have groups who differ only in the disease type, not experience of chronic disease. This approach has been used successfully in previous reports to examine factors in PD such as depression (Gotham et al., 1986) and personality (Leiva, Galvan, & Matius-Guia, 1996; Menza et al., 1993; Poewe et al.,

1983). All assessments were administered by the author of this thesis. Ethical approval was granted for each patient group from the appropriate hospital ethics committees.

Subjects

Subjects were initially contacted by telephone and an appointment made for the full assessment. All of the patients in the PD sample were current or past patients of the National Hospital for Neurology and Neurosurgery in London. Forty-seven PD patients were initially assessed, but data on two were rejected due to changes in diagnoses discovered later (one progressive supranuclear palsy and the other Lewy body

dementia). Twenty-three of the patients were female, the mean age o f the sample was 66.36 years (range 48-79 years). The sample included a wide range of disease

progression; the mean Hoehn and Yahr stage (described in Appendix C) was 2.6 (range 1-5). The OA patients were all current or past patients of the Department of

Rheumatology, Kings College Hospital in London. Nineteen patients were initially assessed, but data on two were rejected due to possible neurological involvement (one showed signs of undiagnosed parkinsonism, the other was at high risk o f having

ischemic white matter lesions). Twelve of the OA patients were female. The mean age of the sample was 67.29 years (range 49-84 years). There was no significant difference between the two groups for age (t(60)=.37, p=.712) but the OA patients had

significantly fewer years of education (0A =11.18, SD=2.54; PD=13.00, SD=2.85; t(58)=-2.30, p=.025).

Assessments

The assessment of the two groups was identical with the exception that Hoehn and Yahr and Schwab and England (Schwab & England, 1969) disease progression stagings were

only appropriate for the PD sample. Activities of daily living (ADL) scores were recorded with a scale previously developed for use in comparing PD and arthritis patients. This 24 point self-report Likert scale assesses aspects of day-to-day life that the patients are physically capable of but not what they actually choose to do. Patients are asked to rate their ability to perform a range of common manual dexterity and mobility tasks such as 'Get out of bed' or 'Open tins' without special aids. This scale is described in more detail in Appendix D. Cognitive function was assessed with the Cambridge Examination of Cognition in the Elderly (CAMCOG, Roth, Huppert, Tym, & Mountjoy, 1988) and the Mini-Mental State Examination (MMSE, Folstein &

Folstein, 1975). Both tools have previously been used successfully in evaluating global cognitive function in PD (Hobson & Meara, 1999). In addition, three tests of executive function were included, the Stroop task (Stroop, 1935), verbal fluency (FAS, Benton,

1986) and the Wisconsin Card Sort Test-short version (WCST, Nelson, 1976). Further details of the cognitive assessments used including scoring methods are given in Appendix E.

In order to assess current psychiatric status the Beck Depression Inventory (BDI, Beck, Ward, Mendelson, Mock, & Erbaugh, 1961), the Hospital Anxiety and Depression Scale (HADS, Zigmond & Snaith, 1983) and the Snaith-Hamilton Pleasure Scale (SHPS, Snaith et al., 1995) were included to give estimates of depression, anxiety and anhedonia respectively.

For the assesment of personality the Tridimensional Personality Questionnaire (TPQ) was used (Cloninger, 1987). This measures three personality dimensions. Novelty Seeking (NS), Harm Avoidance (HA) and Reward Dependence (RD). However, more recent factor analysis has indicated that a previous sub-factor o f RD should be

considered separately (Stallings, Hewitt, Cloninger, Heath, & Eaves, 1996). This fourth factor is called Persistence (P).

Additionally, to assess apathy all patients completed the Self-report version of the Apathy Evaluation Scale (AES-S) and were questioned to obtain scores on the

Researcher rated Apathy Evaluation Scale (AES-R, Marin et al., 1991). The AES-R is administered as a semi-structured interview. Additionally, if a caregiver was available they were asked to complete the Informant version (AES-I). Details of the items and structure of the AES-S, AES-R and AES-I are given in Appendix B.

Administration

A few days before the date of the appointment, all patients were sent copies of the TPQ, BDI and SHPS and were asked to complete these before the researcher arrived, this was performed to minimise the time spent in a single session and reduce the risk of

assessments being aborted due to fatigue. At the beginning of the actual session individual details such as date of birth and gender were collected first, followed by the ADL assesment, and where appropriate, the Hoehn and Yahr and Schwab and England assessments. Following this all patients completed the verbal fluency (FAS) task followed by the CAMCOG examination, which includes the MMSE and category fluency assessments. Next, all patients were asked to complete the HADS and AES-S. If a caregiver was available and the patient consented, the caregiver was asked if they would complete the AES-I. The patients were not allowed to see the caregivers

responses. Following this, the AES-R interview was completed followed by the Stroop task and the WCST. Due to fatigue, time constraints and occasional objections, some assessments were not completed by all patients. The lowest completion rates by the patients were for the HADS and WCST which were only completed by 52 (84%) of the

patients. A fiill list of completion rates is given in Appendix F. Caregiver apathy ratings (AES-I) were only obtained for 30 (66.7%) of the PD patients and five (29%) of the OA patients. The time to complete the full session took approximately two hours per patient.

The AES-R is the most in-depth tool currently available for the assessment of apathy of subjects ‘in the field’ and has proven reliability and validity (discussed in Chapter 2). In addition to this measure, a large amount of data on disability, cognitive and psychiatric status and personality were also collected. This gave the opportunity to examine the contribution of multiple factors to the presence of apathy in PD patients. These factors are considered individually below.

Results of Disability and Apathy

In order to check for a difference in the functional ability of the two patient groups, a t- test was performed on the ADL data. This revealed that there was a small difference between the PD group and the OA group on the ADL scale, with the PD group being slightly more disabled, but this did not reach significance (PD mean=53.45, SD=21.35; OA mean=43.71, SD=16.5; t(59)=-1.69, p=.096). In terms o f apathy, the OA group scored an average of 23.29 (SD=3.82) on the AES-R. This is significantly lower than the PD group who scored 35.29 (SD=11.37), t(60)=-4.24, p<.001. Similar significant group differences were also found using the self-report AES-S version. The PD patients scored 32.95 (SD=9.03) and the OA patients scored 26.88 (SD=5.01), t(59)=-2.61, p=.011. Between group comparisons were not attempted using the informant version (AES-I) given to caregivers. This was because although data was available on 30 PD patients, it was only available on five OA patients. It was considered that the sample size in the latter group was too small to be representative.

In order to assess the consistency of the three measures of apathy, their level of agreement was compared using pairwise parametric correlations. It was found that all pairs were significantly correlated at the level of p<.OOI. The self-report AES-S was positively correlated with the AES-R (.717) and with the AES-I (.766). The highest correlation was between the AES-I and AES-R (.777). To check reliability, 20 PD patients completed the AES-S at a second session three to six months after the initial assessment. The correlation between the two administrations was high, r=.848, p<.001. Both the researcher rated and self-rated apathy scores significantly identified the

differences between the two groups in apathy and are highly correlated. As Marin et al. (1991) have previously shown that the AES-R has greater predictive validity, further analysis was limited to this more in-depth measure.

The distribution of AES-R scores by the OA and PD groups are shown in Figure 3. In the PD sample. It can be seen that there was a bimodal distribution. A cut-point of below 38 identified one peak and a score of 38 or higher the other. This criterion was used to divide the PD patients into two groups. The high apathy PD group (PD-HA) contained 15 patients (33.3%) and the low apathy PD group (PD-LA) contained 30 patients (66.7%).

Using this grouping, none of the OA patients was within the range o f scores in the PD- HA group (38-62). Twenty-five (83.3%) of the PD-LA sample were within the range of scores in the OA group (18-33). If the OA group are compared with the results found using the AES-R with an American sample of healthy individuals of a comparable age,